The architecture really is a huge leap forward, according to people who’ve gone through it in-depth. Interestingly, the huge focus for Fermi is GPU computing. The first actual goods coming out using Fermi should be the GT300 series cards, which, besides the 512 cores sorted into 16 streaming processors with 32 cores each, uses a brand new GDDR5 memory setup.

PC Perspective has an epic write-up breaking down Fermi in detail that’s worth a whirl, and of course Nvidia’s got lots of fluff themselves all about Fermi. Strangely, they don’t explain the name, which sounds like a sad little poodle. [Nvidia, PC Perspective]

They Call It “Open” For a Reason
One of the more excellent aspects of Snow Leopard, actually, is its full-scale deployment of OpenCL 1.0—Open Computing Language—a framework that allows programmers to more easily utilise the full power of mixes of different kinds of processors like GPUs and multi-core CPUs. (Much of the excitement for that is in leveraging the GPU for non-graphical applications.)

OpenCL lives up to its name: It is a royalty-free open standard managed by the Khronos Group, and supported by AMD/ATI, Apple, ARM, IBM, Intel, Nvidia, among others. Interesting thing about this open industry standard is that it was developed and proposed by…Apple.

What Is a Standard?
By “standard”, we’re talking about a format, interface or programming tool that a bunch of companies or people or organisations agree is the way something’s going to get done, whether it’s how a movie is encoded or the way websites are programmed. Otherwise, nothing works. A video that plays on one computer won’t play on another, web sites that work in one browser don’t work in another, etc. With increased connectedness between different machines and different platforms, standards are increasingly vital to progress.

Standards can range from open (anybody can use them, for free) to open with conditions (anybody can use them as long they follow conditions X, Y and Z) to closed (you gotta have permission, and most likely, pay for it). Some companies view standards strictly as royalty machines; others don’t make much money on them, instead using them to make sure developers do things the way they want them to. Apple falls into this latter category, by choice or possibly just by fate.

Kicking the Big Guy in the Shins
Of course, OpenCL isn’t the only open standard that Apple’s had a hand in creating or supporting that actually went industry-wide. When you’re the little guy—as Apple was, and still is in computer OS marketshare, with under 10 percent—having a hand in larger industry standards is important. It keeps your platform and programming goals from getting steamrolled by, say, the de facto “standards” enforced by the bigger guy who grips 90 per cent of the market.

If you succeed in creating a standard, you’re making everybody else do things the way you want them done, but presumably without strong-arming. If you’re doubting how important standards are, look no further than Sony throwing a new one at the wall every week hoping it’ll stick. Or Microsoft getting basically everybody but iTunes to use its PlaysForSure DRM a couple years ago. Or its alternative codecs and formats for basically every genuine industry standard out there. To be sure, there is money to be made in standards, but only if the standard is adopted—and royalties can be collected.

Web Standards: The Big Headache
The web has always been a sore spot in the standards debate. The web is a “universal OS,” or whatever the cloud-crazy pundits call it, but what shapes your experience is your browser and in part, how compliant it is with the tools web developers use to build their products. Internet Exploder shit all over standards for years, and web programmers still want IE6 to die in a fiery eternal abyss.

Enter WebKit, an open source browser engine developed by Apple based off of the KHTML engine. It’s so standards-compliant it tied with Opera’s Presto engine to be the first to pass the Acid3 test. What’s most striking about WebKit isn’t the fact it powers Safari and Google Chrome on the desktop, but basically every full-fledged smartphone browser: iPhone, Android, Palm Pre, Symbian and (probably) BlackBerry. So WebKit hasn’t just driven web standards through its strict adherence to them, but it has essentially defined, for now, the way the “real internet” is viewed on mobile devices. All of the crazy cool web programming you see now made is made possible by standards-compliant browsers.

True, OpenCL and WebKit are open source—Apple’s been clever about the way it uses open source, look no further than the guts of OS X—but Apple is hardly devoted to the whole “free and open” thing, even when it comes to web standards.

All the AV Codecs You Can Eat
The recent debate over video in the next web standards, known collectively as HTML5, shows that: Mozilla supports the open-source Ogg Theora video codec, but Apple says it’s too crappy to become the web’s default video standard—freeing everyone from the tyranny of Adobe’s Flash. Apple says Ogg’s quality and hardware acceleration support don’t match up to the Apple-supported MPEG-4 standardised H.264 codec, which is tied up by licence issues that keep it from being freely distributed and open. (Google is playing it up the middle for the moment: While it has doubts about the performance of Ogg Theora, Chrome has built-in support for it and H.264.)

Apple has actually always been a booster of MPEG’s H.264 codec, which is the default video format supported by the iPhone—part of the reason YouTube re-encoded all of its videos, actually—and gets hardware acceleration in QuickTime X with Snow Leopard. H.264 is basically becoming the video codec (it’s in Blu-ray, people use it for streaming, etc.).

Why would Apple care? It means Microsoft’s WMV didn’t become the leading standard.

A sorta similar story with AAC, another MPEG standard. It’s actually the successor to MP3, with better compression quality—and no royalties—but Apple had the largest role in making it mainstream by making it their preferred audio format for the iPod and iTunes Store. (It saw some limited use in portables a little earlier, but it didn’t become basically mandatory for audio players to support it until after the iPod.) Another bonus, besides AAC’s superiority to MP3: Microsoft’s WMA, though popular for a while, never took over.

FireWire I Mean iLINK I Mean IEEE 1394
Speaking of the early days of the iPod, we can’t leave out FireWire, aka IEEE 1394. Like OpenCL, Apple did a lot of the initial development work (Sony, IBM and others did a lot of work on it as well), presented it to a larger standards body—the Institute of Electrical and Electronics Engineers—and it became the basis for a standard. They tried to charge a royalty for it at first, but that didn’t work out. It’s a successful standard in a lot of ways—I mean, it is still on a lot of stuff like hard drives and camcorders still—but USB has turned out to be more universal, despite being technically inferior. (At least until USB 3.0 comes out, hooray!)

Update: Oops, forgot Mini DisplayPort, Apple’s shrunken take on DisplayPort—a royalty-free video interface standard from VESA that’s also notably supported by Dell—which’ll be part of the official DisplayPort 1.2 spec. Apple licenses it for no fee, unless you sue Apple for patent infringement, which is a liiiiittle dicey. (On the other hand, we don’t see it going too far as industry standard, which is why we forgot about it.)

That’s just a relatively quick overview of some of the standards Apple’s had a hand in one way or another, but it should give you an idea about how important standards are, and how a company with a relatively small marketshare (at least, in certain markets) can use them wield a lot of influence over a much broader domain.

Shaping standards isn’t always for royalty checks or dominance—Apple’s position doesn’t allow them to be particularly greedy when it comes to determining how you watch stuff or browse the internet. They’ve actually made things better for the sake of making things better, at least so far. One glance at the iPhone app approval process should give anybody who thinks they’re the most gracious tech company around reason to watch out for ulterior motives.

Still something you wanna know? Send questions about standards, things that are open other than your mum’s legs or Sony Ultra Memory Stick XC Duo Quadro Micro Pro II to tips@gizmodo.com, with “Giz Explains” in the subject line.

How many of you brushed off NVidia’s Tegra platform announcement a few weeks ago as just another boring processor development? Yeah, thats what I thought. Well, watch this video and you will very quickly change your mind.

The augmented reality game was developed by some boffins at Virginia Georgia Tech. All I have to say is the sooner we can incorporate skittles into gaming, the better place the world will be for everybody. Plus, Optimus Prime aside, this is the first augmented reality application I can actually see taking off.

At the moment, it’s all just concept – there isn’t an actual device out there that’s capable of running it yet. But once Tegra actually launches in devices, hopefully we’ll start seeing some real breakthrough games on mobile devices…

[ARhrrrr! via Wired]

Engadget says details are scant, but mention that AMD says that new DX11 features, such as the Compute Shader, will help improve Windows 7 performance, among other things. They expect their DirectX 11 products to be available in late 2009. [AMD via Engadget]

Along with its neato Moblin 2.0 netbook OS, Intel’s also laid in more detail out what the next generation of Atom looks like—Pineview integrates the CPU, GPU and memory controller onto one chip, making the platform cheaper and less power-hungry overall.

As Ars’ Jon Stokes points out, this is new territory for Intel that “arguably pushes Atom into SoC territory.” It has some implications for Nvidia’s Ion platform. Not only are Intel’s graphics built into the Atom CPU, but Nvidia’s probably going to have major problems from a price standpoint, since Intel can drop the cost of the Atom platform (which it sells for $US25) down even further. Unbundling the Atom CPU—like to pair it with Nvidia’s 9400m for the Ion platform—costs $US45.

Never a dull moment in netbook land. [Ars]

No, I didn’t stutter: GPGPU—general-purpose computing on graphics processor units—is what’s going to bring hot screaming gaming GPUs to the mainstream, with Windows 7 and Snow Leopard. Finally, everbody’s face melts! Here’s how.

What a Difference a Letter Makes
GPU sounds—and looks—a lot like CPU, but they’re pretty different, and not just ’cause dedicated GPUs like the Radeon HD 4870 here can be massive. GPU stands for graphics processing unit, while CPU stands for central processing unit. Spelled out, you can already see the big differences between the two, but it takes some experts from Nvidia and AMD/ATI to get to the heart of what makes them so distinct.

Traditionally, a GPU does basically one thing, speed up the processing of image data that you end up seeing on your screen. As AMD Stream Computing Director Patricia Harrell told me, they’re essentially chains of special purpose hardware designed to accelerate each stage of the geometry pipeline, the process of matching image data or a computer model to the pixels on your screen.

GPUs have a pretty long history—you could go all the way back to the Commodore Amiga, if you wanted to—but we’re going to stick to the fairly present. That is, the last 10 years, when Nvidia’s Sanford Russell says GPUs starting adding cores to distribute the workload across multiple cores. See, graphics calculations—the calculations needed to figure out what pixels to display your screen as you snipe someone’s head off in Team Fortress 2—are particularly suited to being handled in parallel.

An example Nvidia’s Russell gave to think about the difference between a traditional CPU and a GPU is this: If you were looking for a word in a book, and handed the task to a CPU, it would start at page 1 and read it all the way to the end, because it’s a “serial” processor. It would be fast, but would take time because it has to go in order. A GPU, which is a “parallel” processor, “would tear [the book] into a thousand pieces” and read it all at the same time. Even if each individual word is read more slowly, the book may be read in its entirety quicker, because words are read simultaneously.

All those cores in a GPU—800 stream processors in ATI’s Radeon 4870—make it really good at performing the same calculation over and over on a whole bunch of data. (Hence a common GPU spec is flops, or floating point operations per second, measured in current hardware in terms of gigaflops and teraflops.) The general-purpose CPU is better at some stuff though, as AMD’s Harrell said: general programming, accessing memory randomly, executing steps in order, everyday stuff. It’s true, though, that CPUs are sprouting cores, looking more and more like GPUs in some respects, as retiring Intel Chairman Craig Barrett told me.

Explosions Are Cool, But Where’s the General Part?
Okay, so the thing about parallel processing—using tons of cores to break stuff up and crunch it all at once—is that applications have to be programmed to take advantage of it. It’s not easy, which is why Intel at this point hires more software engineers than hardware ones. So even if the hardware’s there, you still need the software to get there, and it’s a whole different kind of programming.

Which brings us to OpenCL (Open Computing Language) and, to a lesser extent, CUDA. They’re frameworks that make it way easier to use graphics cards for kinds of computing that aren’t related to making zombie guts fly in Left 4 Dead. OpenCL is the “open standard for parallel programming of heterogeneous systems” standardised by the Khronos Group—AMD, Apple, IBM, Intel, Nvidia, Samsung and a bunch of others are involved, so it’s pretty much an industry-wide thing. In semi-English, it’s a cross-platform standard for parallel programming across different kinds of hardware—using both CPU and GPU—that anyone can use for free. CUDA is Nvidia’s own architecture for parallel programming on its graphics cards.

OpenCL is a big part of Snow Leopard. Windows 7 will use some graphics card acceleration too (though we’re really looking forward to DirectX 11). So graphics card acceleration is going to be a big part of future OSes.

So Uh, What’s It Going to Do for Me?
Parallel processing is pretty great for scientists. But what about those regular people? Does it make their stuff go faster. Not everything, and to start, it’s not going too far from graphics, since that’s still the easiest to parallelize. But converting, decoding and creating videos—stuff you’re probably using now more than you did a couple years ago—will improve dramatically soon. Say bye-bye 20-minute renders. Ditto for image editing; there’ll be less waiting for effects to propagate with giant images (Photoshop CS4 already uses GPU acceleration). In gaming, beyond straight-up graphical improvements, physics engines can get more complicated and realistic.

If you’re just Twittering or checking email, no, GPGPU computing is not going to melt your stone-cold face. But anyone with anything cool on their computer is going to feel the melt eventually.

Occupying a space only millimeters wide, and supporting the Open GL ES 2.0 standard, Product Manager Remi Pedersen says that the Mali 200, and eventually, Mali 400, are designed to work in a phone that goes 2-3 days between charges. Pedersen says the first phones will appear at the end of 2009, followed by an influx of devices in 2010.

Graphically, games shown were on the level of PS2 and Xbox, able to push a decent number of pixels with a smooth framerate. A port of the original Project Gotham Racing runs on the Mali 200 GPU with virtually no lag and a decent number of polygons.

But they can also provide hardware acceleration for device UIs, process HD video and make Flash usable on mobile devices. ARM says that features such as HD video encode/decode and Flash decoding will be centered more around the multicore Mali 400, which will appear en masse sometime in 2010. And by all accounts, we can probably look forward to seeing this line of Mali GPUs in future netbooks and MIDs.

Here’s quick vid of the Mali 200 in action. It’s pretty smooth for a mobile GPU.

Mobile GPU drivers were so scarce because those drivers are very customised, from hotkeys to smooth suspend and resume functions on each individual laptop. As a result, many advanced users have had to scour forums for drivers hacked together by other users, and that’s not the most reliable way to upgrade or update software. So it’s great to see Nvidia offering broad support for driver updates on their own site, especially since they say they’ve been hard at work to ensure compatibility and retention of features across all manufacturers. Hopefully these official releases will be more stable than the unofficial hacks. [Notebook Review via Engadget]

Nvidia calls its new platform—Atom + GeForce 9400M (the same chipset as the aluminium MacBooks—Ion. Performance-wise, Nvidia is promising 5x faster graphics and 10x faster video transcoding than a standard Atom-powered netbook running on Intel’s current platform. The Call of Duty thing, I still want to see before I totally believe it, but they promised 25-30s running at 1024×768 resolution. Not amazing but playable. Before you ask, it’ll run Crysis, though the results would make you hurt—which is still better than the current netbook crop. (If you proceed to ask anyway or if it will blend, I will ban you.)

So, it’ll run graphics faster, better, meaning netbooks that won’t cry when it comes to video tasks, like playing 1080p Full HD video. And you’ll see more performance benefits as OSes and apps take advantage of GPU acceleration—like Windows 7, Snow Leopard (Hackintosh power!) and any other CUDA or OpenCL app (admittedly not super common yet). Yeah, you’ll be able to actually run Vista Premium and Windows 7 and not hate life (if you’ve got the RAM too, anyway). Supposedly all with comparable battery life to current systems.

Bad news? We won’t see Ion netbooks until about midway through 2009, and when we do they’re going to cost a bit more than other netbooks—”within $US50″ of standard netbook pricing, since Nvidia is positioning them as “premium” netbooks, whatever the hell that means. And this still doesn’t help netbooks’ other serious shortcomings, like multitasking or crummy keyboards.

But at least they’ll suck just a little bit less. [Nvidia]